“With riboswitches, metabolite sensing is part of the messenger RNA, and proteins are not required to control genes.”

Our paper provided the first biochemical proof that messenger RNAs can form structures called "riboswitches" that directly bind metabolites, without the aid of proteins, and subsequently control gene expression. Riboswitches are a simpler form of genetic control and serve as objects of research for those who study how modern gene control systems work, for those who engineer RNA to exhibit novel functions, and for those who are testing theories about how complex life forms emerged from primitive biochemicals.

There was some published (and much unpublished) speculation that RNA molecules could serve as metabolite-binding gene control elements spanning the last decade. In addition, there were data being published as much as 30 years ago that we can now recognize as implicating the existence of riboswitches. What our work has done is provide the convincing evidence that riboswitches exist, that they form binding pockets for specific and high-affinity binding of their target compounds, and that they use defined allosteric mechanisms for converting a ligand-binding event into a gene control output.

We had been using a process called "test tube evolution" to create RNA switches in the laboratory. From trillions of molecular misfits, we isolate rare molecules that function as metabolite-dependent switches in a process that simulates Darwinian evolution. While generating a series of such RNA switches to create an "RNA array" biosensor, we recognized that RNA switches were relatively easy to make, and most likely nature would have made use of this same technology in modern cells, and perhaps in ancient cells from billions of years ago. We speculated that, if riboswitches were in modern organisms, hints of such systems would be in the literature. To date, seven distinct classes of riboswitches have been confirmed by examining gene control mysteries that had been reported by others.

Previously, it was believed that proteins were needed to sense important chemical compounds. Therefore proteins would be required to bind metabolites and subsequently modulate the expression of genes. With riboswitches, metabolite sensing is part of the messenger RNA, and proteins are not required to control genes. For cell biologists, this provides a newfound way in which genes can be turned on or off. For those who seek to understand how life evolved from primitive to complex cells, this finding explains how simple organisms could control cellular process even before proteins emerged in evolution.